The results strongly suggest that the methodological design of the EPR policy has an impact on the intended outcomes, i.e. the medicine prices in the own country. In particular, the simulations run in this study point to considerable relevance of some of the parameters of the EPR methodology. Thus, the findings add to studies that suggested EPR’s ability to reduce medicine prices (19-24).
It has been highlighted in literature (28, 35-37) and in policy debate that the cost-containment capacity of EPR is strongly impaired by referencing to ‘fake prices’ (38) since the real prices are not known due to the confidential character of discounts and managed-entry agreements. Both scenarios of this study that considered discounts confirmed the loss of savings opportunities given the non-consideration of discounted prices in other countries. Even the scenario that only took into account statutory (thus published) discounts showed important price-reducing potentials. As the consideration of published discounts would not imply any breach of confidentiality, an EPR-applying country could implement it rather swiftly. In fact, since its medicine pricing reform of 2017 price data reduced by statutory discounts are taken into consideration in Austria (39). Another technical option for governments to account for discounts could be to follow the example of scenario ‘statutory and commercial discounts’ and to assume a ‘reasonable’ discount. At political level, a debate on price transparency is ongoing, as evidenced by the ‘WHO Transparency Resolution’ adopted by the World Health Assembly in May 2019. This WHO Resolution calls for disclosure of net prices as well as research and development costs for medicines and vaccines (40).
It can be expected that the approach to determine the EPR benchmark price by referring to the lowest priced reference country will lead to lower prices compared to other methodologies. This was, not surprisingly, confirmed by the simulations results. The base case scenario included some European countries that calculated their reference price based on the average of the prices in the reference countries, and particularly for these countries, major decreases compared to the base case were seen. However, it can be discussed whether, or not, a policy of a ‘race to the bottom’ is an intended objective of EPR that is a pricing policy usually applied for new medicines. As an alternative, opportunities for savings could also be achieved from off-patent medicines, as evidence on the price-reducing character of generic competition (41-45) and of tendering (46-50) is available.
The selection of the reference countries is a key decision point in the design of EPR. The WHO Guideline on Country Pharmaceutical Pricing Policies recommends choosing reference countries based on a set of explicitly stated factors (17). With regard to the reference countries, policy-makers have to take two major choices: the size of the basket and the countries to include. With regard to the latter, it is common sense that a focus on lower-priced reference countries will eventually lead to lower prices. There is, however, the risk that particularly in the beginning price setting might be difficult because medicines might not have a price and be marketed in lower-priced countries due to strategic launches of pharmaceutical companies in response to the widespread use of EPR (32, 51, 52). Thus, countries, particularly those referring to lower-priced countries, are advised to have a mechanism in place which allows setting the price even with the product being marketed in a few countries (e.g. alternative countries) and provides for regular price reviews to benefit from price decreases in the reference countries over the years (see also findings of scenario ‘bi-annual revisions’) (18). As a related aspect, it has to be decided whether, or not, there is a need to have large country baskets. This has also to be seen in connection with the resources required for surveying medicine price data to perform EPR, which can be substantial in case of large country baskets (53). In any case, the study findings confirm the importance of a strategic selection of the reference countries: a well-chosen small country basket is not only less resource-intensive but may also achieve lower prices. Indeed, the simulations showed that most countries would pay higher prices (increases by 20% in several cases) if they used a larger basket (scenario ‘large basket’ with the assumption of 30 reference countries).
EPR has been criticised for failing to deliver equity since it does not consider the different income levels of the reference countries (27, 54). As a solution, differential pricing – a policy in which medicine prices are set in line with the countries’ economic status – has been proposed (55-58). Usually, differential pricing and EPR are considered as mutually exclusive policy options. However, in the authors’ perception, this is not necessarily the case. For instance, the prices in the reference countries could be weighted by indicators that reflect the economic situation of these countries (e.g. gross domestic product, PPP). The simulation scenario that was run on PPP-adjusted prices showed lower prices for lower-income countries but also an increased burden due to higher prices for higher-income countries. While accounting for countries’ income would contribute to more equity and fairness, such an approach may still be politically acceptable for high-income countries that are meanwhile also struggling to afford medicines.
The authors acknowledge that the study has limitations. The simulations model had to be based on assumptions (e.g. on the starting price and the kick-off countries), which were simplified compared to the far more complex reality. It was decided to focus on pricing of medicines in the outpatient sector, which resulted in the exclusion of Denmark (EPR only applicable for some hospital medicines). EPR was assumed to be the sole pricing policy in the EPR-applying countries. In the studied countries, the EPR policy is not always used for all medicines but other pricing policies are also applied. For instance, internal price referencing, which considers the prices of comparable medicines (e.g. of the same active ingredient) in the same country, is also commonly applied, in particular for medicines in the off-patent market (e.g. generics, biosimilars, and originator medicines whose patent has expired). Even if EPR is applied, some countries accompany its use by further policies; e.g. price negotiations, for which the EPR-determined benchmark price rather serves as background information.
No scenario in which the price data were weighted by volume was run as these data were not available. Governments frequently lack consumption data of other countries. Weighting by volume data of the own country could be applied as a back-up option, but even these data are not always accessible (e.g. in countries with a fragmented health care system, such as Austria and Germany, aggregated consumption data for the hospital sector are missing).
Medicine prices were the sole outcome indicator of the study. Pharmaceutical prices are indeed a major contributing factor for affordable and equitable access to medicines (2) because in solidarity-based systems (as those of the studied countries) lower prices allow the public payers to treat more patients for the same amount spent. But pharmaceutical expenditure is also influenced by volume. Thus, even if prices were decreased, expenditure may grow as a result of increases in consumption (19). The latter may be attributable to over-use or inappropriate use and also to adjustments of previous under-use. Furthermore, public pharmaceutical expenditure may be lower since the published list prices based on EPR are not the reimbursement prices; confidential discounts reduce the price that the public payers actually pay.
Analysing medicine prices as outcome parameter, the study focused on the cost-containment potential of EPR. Thus, it did not consider further objectives that policy-makers may aim to achieve, e.g. to facilitate timely access to medicines, to support the local industry, to incentivise research-oriented pharmaceutical industry to invest into research and development or to ensure that the same price for a medicine is charged in all pharmacies throughout the whole country. Other pharmaceutical (pricing) policies might be more appropriate to reach these objectives.
Finally, the study only analysed the impact of the EPR methodology on the prices in the same countries. EPR is known for its spill-over effects on access in other countries (e.g. launch delays in lower-income countries with lower medicine prices) (11, 51, 52) but this was not scope of this research.
Despite its limitations, the research has important policy implications: It reminds policy-makers to carefully take methodological choices when they implement EPR. It is not simply a question of whether, or not, they apply EPR, but also how they do so. Some parameters of the EPR methodology generate higher cost-containment impacts than others, and if pricing authorities and public payers aim to achieve lower prices through EPR, specific parameters, i.e. consideration of discounts and lowest prices as benchmark, appear to be most appropriate ones. The first, however, would imply a major change in the design of most EPR policies in place, since, as shown in the survey, only very few countries consider mandatory discounts which are publicly accessible. Changing EPR legislation by indicating discounted prices as reference would additionally signal the interest of governments to overcome the current challenge of ‘fake prices’ (38).